[Qemu-devel] [PATCH] virtio: guard against negative vq notifies

[Qemu-devel] [PATCH] virtio: guard against negative vq notifies

[Stefan Hajnoczi]

The virtio_queue_notify() function checks that the virtqueue number is
less than the maximum number of virtqueues.  A signed comparison is used
but the virtqueue number could be negative if a buggy or malicious guest
is run.  This results in memory accesses outside of the virtqueue array.

It is risky doing input validation in common code instead of at the
guest<->host boundary.  Note that virtio_queue_set_addr(),
virtio_queue_get_addr(), virtio_queue_get_num(), and many other virtio
functions do *not* validate the virtqueue number argument.

Instead of fixing the comparison in virtio_queue_notify(), move the
comparison to the virtio bindings (just like VIRTIO_PCI_QUEUE_SEL) where
we have a uint32_t value and can avoid ever calling into common virtio
code if the virtqueue number is invalid.

Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
---
 hw/syborg_virtio.c |    4 +++-
 hw/virtio-pci.c    |    4 +++-
 hw/virtio.c        |    4 +---
 3 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/hw/syborg_virtio.c b/hw/syborg_virtio.c
index 2f3e6da..00c7be8 100644
--- a/hw/syborg_virtio.c
+++ b/hw/syborg_virtio.c
@@ -146,7 +146,9 @@ static void syborg_virtio_writel(void *opaque, target_phys_addr_t offset,
             vdev->queue_sel = value;
         break;
     case SYBORG_VIRTIO_QUEUE_NOTIFY:
-        virtio_queue_notify(vdev, value);
+        if (value < VIRTIO_PCI_QUEUE_MAX) {
+            virtio_queue_notify(vdev, value);
+        }
         break;
     case SYBORG_VIRTIO_STATUS:
         virtio_set_status(vdev, value & 0xFF);
diff --git a/hw/virtio-pci.c b/hw/virtio-pci.c
index c19629d..6862aa7 100644
--- a/hw/virtio-pci.c
+++ b/hw/virtio-pci.c
@@ -348,7 +348,9 @@ static void virtio_ioport_write(void *opaque, uint32_t addr, uint32_t val)
             vdev->queue_sel = val;
         break;
     case VIRTIO_PCI_QUEUE_NOTIFY:
-        virtio_queue_notify(vdev, val);
+        if (val < VIRTIO_PCI_QUEUE_MAX) {
+            virtio_queue_notify(vdev, val);
+        }
         break;
     case VIRTIO_PCI_STATUS:
         if (!(val & VIRTIO_CONFIG_S_DRIVER_OK)) {
diff --git a/hw/virtio.c b/hw/virtio.c
index 6e8814c..a651860 100644
--- a/hw/virtio.c
+++ b/hw/virtio.c
@@ -585,9 +585,7 @@ void virtio_queue_notify_vq(VirtQueue *vq)
 
 void virtio_queue_notify(VirtIODevice *vdev, int n)
 {
-    if (n < VIRTIO_PCI_QUEUE_MAX) {
-        virtio_queue_notify_vq(&vdev->vq[n]);
-    }
+    virtio_queue_notify_vq(&vdev->vq[n]);
 }
 
 uint16_t virtio_queue_vector(VirtIODevice *vdev, int n)
-- 
1.7.4.4

Re: [Qemu-devel] [PATCH] virtio: guard against negative vq notifies

[Stefan Hajnoczi]

On Sun, May 8, 2011 at 10:29 PM, Stefan Hajnoczi
<stefanha@linux.vnet.ibm.com> wrote:
> The virtio_queue_notify() function checks that the virtqueue number is
> less than the maximum number of virtqueues.  A signed comparison is used
> but the virtqueue number could be negative if a buggy or malicious guest
> is run.  This results in memory accesses outside of the virtqueue array.
>
> It is risky doing input validation in common code instead of at the
> guest<->host boundary.  Note that virtio_queue_set_addr(),
> virtio_queue_get_addr(), virtio_queue_get_num(), and many other virtio
> functions do *not* validate the virtqueue number argument.
>
> Instead of fixing the comparison in virtio_queue_notify(), move the
> comparison to the virtio bindings (just like VIRTIO_PCI_QUEUE_SEL) where
> we have a uint32_t value and can avoid ever calling into common virtio
> code if the virtqueue number is invalid.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
> ---
>  hw/syborg_virtio.c |    4 +++-
>  hw/virtio-pci.c    |    4 +++-
>  hw/virtio.c        |    4 +---
>  3 files changed, 7 insertions(+), 5 deletions(-)

Please consider this patch for the stable tree once Michael is happy:
http://patchwork.ozlabs.org/patch/94604/

Stefan

Re: [Qemu-devel] [PATCH] virtio: guard against negative vq notifies

[Michael S. Tsirkin]

On Sun, May 08, 2011 at 10:29:07PM +0100, Stefan Hajnoczi wrote:
> The virtio_queue_notify() function checks that the virtqueue number is
> less than the maximum number of virtqueues.  A signed comparison is used
> but the virtqueue number could be negative if a buggy or malicious guest
> is run.  This results in memory accesses outside of the virtqueue array.
> 
> It is risky doing input validation in common code instead of at the
> guest<->host boundary.  Note that virtio_queue_set_addr(),
> virtio_queue_get_addr(), virtio_queue_get_num(), and many other virtio
> functions do *not* validate the virtqueue number argument.
> 
> Instead of fixing the comparison in virtio_queue_notify(), move the
> comparison to the virtio bindings (just like VIRTIO_PCI_QUEUE_SEL) where
> we have a uint32_t value and can avoid ever calling into common virtio
> code if the virtqueue number is invalid.
> 
> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>

Looks good to me. Didn't test but
Acked-by: Michael S. Tsirkin <mst@redhat.com>

Anthony, you are going to merge that? stable trees need that too,
probably very far back ...

> ---
>  hw/syborg_virtio.c |    4 +++-
>  hw/virtio-pci.c    |    4 +++-
>  hw/virtio.c        |    4 +---
>  3 files changed, 7 insertions(+), 5 deletions(-)
> 
> diff --git a/hw/syborg_virtio.c b/hw/syborg_virtio.c
> index 2f3e6da..00c7be8 100644
> --- a/hw/syborg_virtio.c
> +++ b/hw/syborg_virtio.c
> @@ -146,7 +146,9 @@ static void syborg_virtio_writel(void *opaque, target_phys_addr_t offset,
>              vdev->queue_sel = value;
>          break;
>      case SYBORG_VIRTIO_QUEUE_NOTIFY:
> -        virtio_queue_notify(vdev, value);
> +        if (value < VIRTIO_PCI_QUEUE_MAX) {
> +            virtio_queue_notify(vdev, value);
> +        }
>          break;
>      case SYBORG_VIRTIO_STATUS:
>          virtio_set_status(vdev, value & 0xFF);
> diff --git a/hw/virtio-pci.c b/hw/virtio-pci.c
> index c19629d..6862aa7 100644
> --- a/hw/virtio-pci.c
> +++ b/hw/virtio-pci.c
> @@ -348,7 +348,9 @@ static void virtio_ioport_write(void *opaque, uint32_t addr, uint32_t val)
>              vdev->queue_sel = val;
>          break;
>      case VIRTIO_PCI_QUEUE_NOTIFY:
> -        virtio_queue_notify(vdev, val);
> +        if (val < VIRTIO_PCI_QUEUE_MAX) {
> +            virtio_queue_notify(vdev, val);
> +        }
>          break;
>      case VIRTIO_PCI_STATUS:
>          if (!(val & VIRTIO_CONFIG_S_DRIVER_OK)) {
> diff --git a/hw/virtio.c b/hw/virtio.c
> index 6e8814c..a651860 100644
> --- a/hw/virtio.c
> +++ b/hw/virtio.c
> @@ -585,9 +585,7 @@ void virtio_queue_notify_vq(VirtQueue *vq)
>  
>  void virtio_queue_notify(VirtIODevice *vdev, int n)
>  {
> -    if (n < VIRTIO_PCI_QUEUE_MAX) {
> -        virtio_queue_notify_vq(&vdev->vq[n]);
> -    }
> +    virtio_queue_notify_vq(&vdev->vq[n]);
>  }
>  
>  uint16_t virtio_queue_vector(VirtIODevice *vdev, int n)
> -- 
> 1.7.4.4